00:00 |
signpost |
trivially, all programs could also be represented as lookup tables from inputs to outputs. |
00:00 |
signpost |
this is called caching. what's different in your usage? |
00:00 |
verisimilitude |
Similarly, my Words program won't parse its input; it will use a trie of every possible word. |
00:00 |
verisimilitude |
Mine can be trivially proven correct, signpost. |
00:01 |
verisimilitude |
This also makes it easier to implement such programs in, say, machine code, as there's less to do. |
00:01 |
signpost |
yes, if you write the correct answers for every possible input in a lookup table, your output will be correct, lol. |
| |
↖ |
00:11 |
verisimilitude |
Yes and, through composition rules, those tables can be compressed at the expense of a little extra computation. |
00:12 |
verisimilitude |
A finite table may then be stretched unto infinity. |
| |
↖ |
| |
~ 31 minutes ~ |
00:44 |
asciilifeform |
http://logs.nosuchlabs.com/log/asciilifeform/2022-04-21#1097904 << funnily enuff, this kinda thing was rampant in the 8bit micro days. and not only for e.g. trig functions, but sometimes even (typically 4 bit) multiplication table. faster than 'egypt'... |
00:44 |
dulapbot |
Logged on 2022-04-21 20:00:38 signpost: yes, if you write the correct answers for every possible input in a lookup table, your output will be correct, lol. |
00:46 |
asciilifeform |
for extra laffs, sometimes the table entries were ~not~ correct |
00:52 |
verisimilitude |
Yes; Intel comes to mind. |
| |
~ 35 minutes ~ |
01:28 |
verisimilitude |
http://logs.nosuchlabs.com/log/asciilifeform/2022-04-21#1097894 To walk around and to walk through are the meanings. What was meant by them in particular? |
01:28 |
dulapbot |
Logged on 2022-04-21 19:42:54 signpost: I know what the parts of those two words are and mean; do you know what I meant by making reference to them? |
| |
~ 15 minutes ~ |
01:43 |
signpost |
asciilifeform: sure, and before this, big books of them. |
01:43 |
signpost |
caching's a fine thing but I'm curious why verisimilitude is presenting it as a revolution in computing. |
01:44 |
signpost |
verisimilitude: regarding the latin words, they're conceptually related. would your representation indicate this relatedness? |
01:47 |
verisimilitude |
This isn't caching. |
01:47 |
verisimilitude |
Caching is used to confuse performance characteristics. This makes them certain. |
01:48 |
verisimilitude |
Also, people don't tend to try to apply this to an entire program, only pieces. |
01:48 |
verisimilitude |
Yes it would, signpost. |
01:49 |
verisimilitude |
For Latin, I like the idea of allowing implicit concatenation, meaning I like the idea of allowing ``no space'' as one possibility between words. |
01:50 |
verisimilitude |
For English and other living languages, this would be foolish, but with Latin one can expect a better class of user. |
01:51 |
verisimilitude |
So CIRCVMAMBVLĀRE and PERAMBVLĀRE could be represented as two words, each. |
01:51 |
verisimilitude |
I've already thought of this, signpost. |
01:51 |
signpost |
you're pretty damned bad at explaining things, my dude. |
01:52 |
verisimilitude |
I could link to another discussion, to help. |
01:53 |
verisimilitude |
My system clearly needs to be able to express punctuation between words. |
01:53 |
verisimilitude |
For Latin, one possible punctuation could be ``no space'', to handle exactly these words. |
01:55 |
Mocky |
verisimilitude: lookup tables and machine code and conlangs... are these in service to solve some bigger problem you'd like to solve or more just "A finite wank may be stretched to infinity"? |
01:55 |
dulapbot |
Logged on 2022-04-21 20:11:11 verisimilitude: A finite table may then be stretched unto infinity. |
01:55 |
verisimilitude |
Yes. |
01:56 |
verisimilitude |
I want to program without text, Mocky. |
01:56 |
verisimilitude |
Here is the first article, and an implementation. |
01:57 |
verisimilitude |
However, what passes for text doesn't much suit human languages, either. |
01:57 |
signpost |
http://verisimilitudes.net/2018-06-06 << here you casually use "better", then indicate that perhaps the category in which you deem "better" is storage efficiency? |
01:57 |
verisimilitude |
This is the other end of programming without text. |
01:57 |
signpost |
which if so, sure, you've discovered compression. I think we did this thread. |
01:57 |
verisimilitude |
Good compression requires a high-level model, signpost. |
01:58 |
signpost |
hm, so you're atttempting to construct an abstract model of language which you can manipulate with a non-linguistic interface? |
01:58 |
verisimilitude |
As we've discussed, it's like MIDI, but for human language. |
01:58 |
verisimilitude |
Rephrase that. |
01:58 |
signpost |
no. |
01:58 |
verisimilitude |
Alright. |
01:58 |
signpost |
lol, asking a cogent question |
01:58 |
signpost |
*ask |
01:58 |
Mocky |
Once you can program without text, what benefits will you realize? |
01:58 |
verisimilitude |
I suppose so, signpost. |
01:59 |
verisimilitude |
I type less, eliminate categories of mistakes, get a nicer interface, and other niceties. |
02:00 |
verisimilitude |
These tools will enable me to see more than I otherwise could. |
02:01 |
verisimilitude |
The MMC makes it very easy to reuse machine code as data, which an assembler obscures. |
02:02 |
verisimilitude |
An assembler forces the user to remember mnemonics and can only die when facing errors; my tool does neither. |
02:02 |
verisimilitude |
There are one and two pass assemblers; the MMC is a ``no-pass'' tool. |
02:02 |
signpost |
the unifying vision behind these is not clear in these statements. |
02:02 |
signpost |
not saying there isn't one. |
02:03 |
verisimilitude |
I'll recycle someone else's words. |
02:04 |
verisimilitude |
``I think I see the common thread between the two examples you give: a domain mismatch, where the technology used to manipulate a thing is not made of the same primitives as the thing itself. |
02:04 |
verisimilitude |
A technology that fits its problem domain well can be expected to accomplish two things: first, it should be efficient, by allocating resources only as needed; second, it should disallow elements which are foreign to the domain. |
02:04 |
verisimilitude |
An example of the latter: a program is conceptually made of things like expressions and statements, not characters; if a tool allows typos to be introduced into a program, the tool doesn't fit the domain well.'' |
02:12 |
signpost |
so I've got another "what it's not", wanton use of bare "should" like we know good things when the lord speaks in our hearts, and a third bare assertion that's also a mess. |
02:13 |
signpost |
since I'm feeling charitable, one thing you want to eliminate is being able to give the computer invalid inputs, yes? |
02:14 |
Mocky |
When I look at the volume of your IRC output and the way you drop coupla hundred lines of poetry in front of your UDP binding, and when I imagine how much typing & backspace it must have taken to get dozens of Ada comment lines to all be exactly 100 characters wide, I see that contrary to what you say, your actions suggest that you typing text very highly in the present |
02:15 |
Mocky |
*that you value, rather |
02:17 |
verisimilitude |
Yes, signpost. |
02:17 |
verisimilitude |
I've no choice currently, Mocky. |
02:17 |
verisimilitude |
I value language, certainly, and this is the way available to me. |
02:18 |
verisimilitude |
There can be better ways. |
02:18 |
signpost |
you also wish the computer's state to be as inspectable as possible, thus want programs to (as much as is possible) be tables of enumerated categories of things? |
02:18 |
verisimilitude |
Yes; code is disgusting where it be unnecessary. |
02:19 |
signpost |
I don't care about your aesthetic desires. |
02:19 |
signpost |
but I can see utility in cleanly separating inspectable state from algorithmics |
02:19 |
verisimilitude |
The less code, the more easily it may be manipulated. |
02:19 |
verisimilitude |
How many issues arise from scale alone? |
02:20 |
signpost |
yep. in my example of the two latin words, in my mind the prepositions would be in a table if you like, and the stems another, conjugations yet another. |
| |
↖ |
02:20 |
verisimilitude |
When the entire code be a few kibibytes, efficient replacement algorithms may be elided. |
02:21 |
verisimilitude |
It's not hard to align text so, Mocky; I do that with all of my writings now; view it over Gopher. |
02:21 |
verisimilitude |
Observe: |
02:21 |
verisimilitude |
printf '2022-01-31.read\r\n' | netcat verisimilitudes.net 70 | less |
02:24 |
verisimilitude |
I'm profoundly bored, so perhaps some of this is ``a wank'', but that's irrelevant. |
02:26 |
Mocky |
I'm trying to see how these pieces fit together for you verisimilitude but it just all sounds so crackpot |
02:26 |
signpost |
eh the parts I pulled out above make a bit of sense. |
02:26 |
verisimilitude |
Isn't this channel for crackpots? |
02:26 |
signpost |
I don't grok how verisimilitude wants to pull them into a useful system, but hey. |
02:28 |
signpost |
verisimilitude: did this map onto your idea of it? |
02:28 |
dulapbot |
Logged on 2022-04-21 22:18:49 signpost: yep. in my example of the two latin words, in my mind the prepositions would be in a table if you like, and the stems another, conjugations yet another. |
02:29 |
verisimilitude |
It's close, perhaps. |
02:31 |
verisimilitude |
All words would be in a dictionary; each word would be marked with such information; the declining and conjugating happens at the site of a word's usage. |
02:31 |
signpost |
only one dictionary, not distinguished by part of speech? |
02:31 |
verisimilitude |
It depends. |
02:32 |
verisimilitude |
For AMBVLĀRE, it would be a first grouping verb, and its base AMBVL stored in a separate table. |
02:33 |
verisimilitude |
For IN, that should likely be stored twice, once as a preposition, and once as the negating prefix. |
02:33 |
* |
asciilifeform waits to see what happens when verisimilitude discovers irregular verbs |
02:33 |
verisimilitude |
Those are stored specially. |
02:34 |
verisimilitude |
So, ESSE would likely need to be a table storing every mutation explicitly. |
02:34 |
verisimilitude |
I'm holding off on implementing this until I'm certain I won't run into unforeseen issues, asciilifeform. |
02:35 |
verisimilitude |
I must learn first and model second. |
02:35 |
asciilifeform |
verisimilitude: you aint doing this simply for compactness of encoding, i take it. cuz gzip will prolly still, perhaps frustratingly, win (see if can see why) |
02:36 |
verisimilitude |
No, I ain't. |
02:36 |
verisimilitude |
I view it as the right thing. |
02:37 |
asciilifeform |
verisimilitude: what do you aim to do once you have e.g latin encoder ? |
02:37 |
verisimilitude |
Imagine a learning tool showing ancient Latin, with words coloured and styled by usage, but drawing from this representation; that certainly would beat gzip. |
02:38 |
verisimilitude |
I intend to play with it, for one. |
02:38 |
asciilifeform |
an 'objdump' that eats e.g. caesar's gallic wars ? |
02:38 |
verisimilitude |
I could try my luck bothering unsuspecting Latinists, for another. |
02:38 |
asciilifeform |
a++ |
02:38 |
verisimilitude |
Yes. |
02:39 |
verisimilitude |
I just want it to exist. |
02:39 |
asciilifeform |
fair'nuff |
02:39 |
asciilifeform |
eggog handling for this type of thing imho interesting problem. (texts on the net tend to be ocr product and fulla typos) |
02:39 |
signpost |
were you to say more about what it'll do, and less about your feelings, that'd help the model of the item exist in >1 head. |
02:39 |
verisimilitude |
I try to live my life as if I could die soon; well, I want to die less unfulfilled. |
02:40 |
* |
signpost at least has a better picture. |
02:40 |
signpost |
ah don't worry, you'll always have shit you were going to do the day after you died. |
02:40 |
verisimilitude |
The implementation and a manual would do better than that, signpost. |
02:40 |
signpost |
perhaps. I'll read that when it arises. |
02:41 |
verisimilitude |
I'm glad; I'll try to not disappoint. |
| |
~ 2 hours 40 minutes ~ |
05:21 |
adlai |
yaknow, the analogy of "MIDI for language" captures exactly what I find endlessly irksome about your projects. |
05:24 |
adlai |
don't get me wrong, I do not intend to discourage you. I have a healthy respect for 'diversity', as described in the article linked here, so I recognize that it's important for others to also work on things that don't align perfectly with my preferences. |
05:24 |
dulapbot |
Logged on 2022-04-19 14:45:22 asciilifeform: to revisit upstack thread, before it is lost in digression : what asciilifeform , in the 'if wishes were horses' sense, would ideally be up to, is running a 'Klein type-1 organization'. but impossible w/ avail. resources. so , experimenting with what is possible. |
05:25 |
adlai |
however, MIDI encoding loses so much nuance. |
| |
~ 1 hours 36 minutes ~ |
07:01 |
verisimilitude |
This was asciilifeform's analogy, know. |
07:02 |
verisimilitude |
Compare it to sheet music instead of a recording, say. |
| |
~ 1 hours 3 minutes ~ |
08:05 |
crtdaydreams |
er using keccak v off diana_coman's site and patching for vtools doesn't work, and v itself "cannot find vpatch _____ in /home/$USER/patches" |
08:05 |
crtdaydreams |
dunno if outdated, broken or wat |
08:05 |
crtdaydreams |
patch is there and errythin |
| |
~ 6 hours 58 minutes ~ |
15:03 |
* |
asciilifeform was gonna link crtdaydreams to phf's orig. vtools, which asciilifeform uses, but phf's www appears to be down? |
| |
↖ |
15:07 |
asciilifeform |
meanwhile will link to shinohai's alt-vtron. |
| |
↖ ↖ |
15:09 |
asciilifeform |
$ticker btc usd |
15:09 |
busybot |
Current BTC price in USD: $39818.22 |
15:09 |
asciilifeform |
!w poll |
15:09 |
watchglass |
Polling 15 nodes... |
15:09 |
watchglass |
54.39.156.171:8333 : (ns562940.ip-54-39-156.net) Alive: (0.112s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 |
15:09 |
watchglass |
205.134.172.4:8333 : (172-4.core.ai.net) Alive: (0.113s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 |
15:09 |
watchglass |
71.191.220.241:8333 : (pool-71-191-220-241.washdc.fios.verizon.net) Alive: (0.152s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 (Operator: asciilifeform) |
15:09 |
watchglass |
205.134.172.6:8333 : (172-6.core.ai.net) Alive: (0.081s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=733026 |
15:09 |
watchglass |
205.134.172.26:8333 : Alive: (0.141s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=733026 |
15:09 |
watchglass |
205.134.172.28:8333 : Alive: (0.084s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=733026 (Operator: whaack) |
15:09 |
watchglass |
208.94.240.42:8333 : Alive: (0.213s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 |
15:09 |
watchglass |
205.134.172.27:8333 : Alive: (0.278s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 (Operator: asciilifeform) |
15:09 |
watchglass |
54.38.94.63:8333 : (ns3140226.ip-54-38-94.eu) Alive: (0.259s) V=88888 (/therealbitcoin.org:0.8.88.88/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 |
15:09 |
watchglass |
82.79.58.192:8333 : (static-82-79-58-192.rdsnet.ro) Alive: (0.328s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 |
15:09 |
watchglass |
94.176.238.102:8333 : (2ppf.s.time4vps.cloud) Alive: (0.378s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=732636 |
15:09 |
watchglass |
103.36.92.112:8333 : (terebe.ns01.net) Alive: (0.591s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 |
15:09 |
watchglass |
103.6.212.28:8333 : Alive: (0.732s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=485763 (Operator: whaack) |
15:10 |
watchglass |
75.106.222.93:8333 : Alive: (0.586s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=733026 |
15:10 |
watchglass |
143.202.160.10:8333 : Busy? (No answer in 100 sec.) |
| |
~ 16 minutes ~ |
15:27 |
phf |
http://logs.nosuchlabs.com/log/asciilifeform/2022-04-22#1098024 << what ip was supposed to be that, is that btcbase or the blog? |
15:27 |
dulapbot |
Logged on 2022-04-22 11:02:33 asciilifeform: was gonna link crtdaydreams to phf's orig. vtools, which asciilifeform uses, but phf's www appears to be down? |
15:28 |
asciilifeform |
phf: blog |
15:28 |
phf |
ah, yeah.. it goes down from time to time and i havne't figured it out yet: it's an actual bug in openbsd's httpd |
15:29 |
phf |
some combination of fastcgi pass through and http authorization sections result in a heizenbug |
15:29 |
phf |
also restarted.. |
15:30 |
phf |
i could also do an auto restart :> but i think this way i'll remember to look at the root cause at some point |
| |
~ 22 minutes ~ |
15:53 |
asciilifeform |
hm, still loox dead |
15:53 |
asciilifeform |
phf: interesting bug, maybe exploitable..? |
15:53 |
phf |
asciilifeform: http://barksinthewind.com, new ip also |
15:54 |
asciilifeform |
aa |
15:54 |
asciilifeform |
ty phf |
15:54 |
asciilifeform |
worx! |
15:55 |
* |
asciilifeform fixed link to phf on own www to point to above |
15:55 |
phf |
i should probably post the rename patches, for interested parties |
15:56 |
* |
asciilifeform would read |
15:56 |
asciilifeform |
phf: iirc the rename hack makes for patches that can no longer be applied by hand (patch -p0 < foo... ) tho |
| |
↖ |
15:57 |
phf |
yes, but that's something that should've been brought up before i spent significant amount of time hacking on it! |
15:58 |
phf |
you had that hack to pack everything into a container, then vdiff, but that didn't catch on either |
15:59 |
asciilifeform |
phf himself warned folx re subj |
15:59 |
dulapbot |
(trilema) 2018-10-06 phf: format breaks only in a sense that gnu patch won't press it. current vpatches that don't delete/rename (since the feature is not there) will never the less work with any future changes to vtools |
16:00 |
asciilifeform |
phf: the container thing for ref. |
16:01 |
asciilifeform |
abstracted over the files/dirs aspect. |
16:02 |
asciilifeform |
would've handily solved the rename/move thing, too. but at that pt was tall order, to convince folx to switch formats. |
16:03 |
phf |
in some sense it's not much of an improvement over "patch -p0" not working, since now still need non-stock tool |
16:06 |
asciilifeform |
there yes, defo |
16:07 |
phf |
one thing about vtools, i discovered that gnat has code formatting tool, which produces better result than janky emacs indent. i kind of want to regrind vtools for it, but will obviously also break sigs from people who are no longer with us |
16:09 |
asciilifeform |
a reindent patch aint so hard to read imho |
16:10 |
asciilifeform |
(potentially could even bake a mechanical litmus for 'this patch only reindent?') |
16:10 |
asciilifeform |
.. or simply compare bins locally (if only reindent, will be ==) |
16:24 |
phf |
just wire gnatpp directly into vtools, based on extension. fuck my shit up, fam |
16:25 |
phf |
"what kind of man is vtools man? you know him by his pristine indented ada code." |
16:26 |
* |
asciilifeform has various arguably nonstandard formatting in certain proggies, but afaik no unusual indentation per se |
16:27 |
asciilifeform |
at least not in 0th column, that is |
16:27 |
phf |
asciilifeform: i mostly stopped using emacs (i boot it almost entirely unconfigured to slime) so i started to appreciate standard indent tools for specific languages |
16:28 |
asciilifeform |
phf: what were you using in place of emacs ? (or simply meant stopped editing text?) |
16:30 |
phf |
asciilifeform: really context dependent, but i write text with nvi or ed |
16:32 |
asciilifeform |
phf: not yet 'with telegraph key' but still sportsmanlike, a++ |
16:33 |
phf |
there's that naggum rant against perl approach (i think it's the one where the metaphor is either hacking a road through woods, instead of taking a proper road, because the tool is really good at hacking roads, or smth), and i sort of realized that emacs is exactly perl approach by mass now |
16:34 |
asciilifeform |
subj |
16:34 |
phf |
ty |
16:38 |
phf |
if could somehow reduce the surface to emacs 19, or cmucl's hemlock, but otherwise you have all this hidden state, that kicks in on triggers, and fucks your life up. |
16:39 |
phf |
each new upgrade brings creative modernizations and new dwim behaviors, and deprications, which all somehow not work /quite/ right |
16:39 |
asciilifeform |
imho 19 'peak emacs' |
16:39 |
phf |
missing unicode though, even if you were to figure out how to build it on a present day linux |
16:40 |
phf |
same issue with e.g. hemlock |
16:40 |
* |
asciilifeform not tried the latter |
16:41 |
phf |
until cmucl stopped building on all my systems without some major porting effort i used hemlock as a kind of poor man's lisp machine substrate |
16:42 |
phf |
it had some neat integration features, like buffer's with gray stream interfaces: (format buffer "...") just worked™ |
16:42 |
asciilifeform |
phf: how did it compare to e.g. 'climacs' ? |
16:42 |
phf |
it wasn't written by autists :D |
16:43 |
asciilifeform |
( oblig. cmucl thrd ) |
16:43 |
dulapbot |
(trilema) 2017-12-29 asciilifeform: sbcl is a pretty interesting example of one of those spiked pits, like e.g. gcc -- items to which there is no practical alternative except 'throw away the comp and build log cabin'. but asciilifeform does not have phf's deep historical view of sbcl/cmucl ; asciilifeform arrived into the spiked pit directly |
16:44 |
asciilifeform |
from phf's historical descriptions, sounds interesting and possibly worth reviving tho |
16:44 |
asciilifeform |
( in light of e.g. ) |
16:44 |
dulapbot |
(trilema) 2017-10-08 phf: fwiw all our production lisp runs on sbcl, including btcbase. as much as i'm pimping cmucl, it's not "modern" enough to host a website on unix. i still think it's a better target for a hypothetical on the iron common lisp |
16:45 |
* |
asciilifeform had similar interest in 'kyoto commonlisp' but not got far in |
16:45 |
phf |
cmucl is also significantly scarred, but stills retains much of "lisp machine nature" |
16:47 |
phf |
honestly this is of historical interest mostly, requires much indepth study of lisp lore, to recognize what was there before the micro people fucked things up, and then having done that, does one still have time to revive it |
16:48 |
asciilifeform |
hm scarred in what way ? |
16:49 |
phf |
asciilifeform: opportunistic decisions for the sake of speed or convenience or similar at the expense of adherence to original architectural blueprints |
16:49 |
asciilifeform |
a |
16:50 |
* |
asciilifeform encourages phf to post a bit re subj, atm this is squarely 'buried in the sands' from archaological pov |
16:50 |
asciilifeform |
*archaeological |
16:51 |
* |
asciilifeform found a few bits in old l0gz |
16:51 |
dulapbot |
(trilema) 2016-02-25 phf: i kind of look at it like version 0.5.3. there's barely any development on it, very little actively harmful code, if you don't count asdf 3.* in contribs. it works, it can be studied and improved upon, as it stands it's utterly ignored by the cool kids. there are also of course technical merits, the compiler is conceptually identical and on average almost as good (there were many micro optimizations done by sbcl team, but ev |
16:53 |
* |
asciilifeform was looking through prev. thrds re subj |
16:53 |
asciilifeform |
e.g. |
16:53 |
dulapbot |
(trilema) 2016-06-23 phf: well, native threads were added to sbcl at the expense of good gc |
16:55 |
phf |
this is before i got court marshaled for unstable log due to cmucl issues :> |
16:56 |
asciilifeform |
asciilifeform to this day a bit puzzled wai the heavier, moar complicated 'modernized' sbcl was stable, but the lighter/moar classic, 'carburetted' cmucl (as i understand) suffered hard crashes |
16:57 |
asciilifeform |
goes rather against asciilifeform's usual intuition. |
16:57 |
phf |
asciilifeform: sharp edges |
16:57 |
phf |
they need to be worked around and smoothed |
16:58 |
* |
asciilifeform pictures cmucl as sumthing like 'trb' in re sbcl's 'prb' |
| |
↖ |
16:59 |
phf |
e.g. cmucl friendly web server, araneida, is a collection of hacks more so than a complete software package. i would often come acros snippets posted to forums where some problem was worked around, but for whatever reason never upstreamed. |
17:00 |
asciilifeform |
phf: iirc the crashes were in the runtime per se (i.e. segfault) neh? rather than bug in www lib |
17:00 |
phf |
no, was mostly issue with threads |
17:00 |
phf |
or rather greenlets |
17:00 |
asciilifeform |
a hm. |
17:01 |
phf |
to make things work, i need to stuff greenlet yields in a handful of known long running processes (like e.g. search) |
17:02 |
asciilifeform |
a, so if asciilifeform correctly understands, had to resort in fact to running unthreadsafe coad in threads, to make wwwism go ? |
17:03 |
phf |
no no, you know how greenlets work? "thread" is really just a non-interruptable code, which runs until you either explicitly yield, the code terminates, or you sit on an io |
17:03 |
asciilifeform |
'cooperative multiprocess' aha |
17:04 |
phf |
if you e.g. have a process that takes 3s to run, until yield, means that /nothing/ else works during these 3s. if you have a process that takes 20s to run, then your system starts time outing left and right |
17:04 |
asciilifeform |
( funnily enuff this is effectively how threading in trb worx! ) |
17:04 |
dulapbot |
Logged on 2020-09-25 12:57:05 asciilifeform: whaack: this happens on account of the asinine pseudo-multithreading in trb. i.e. noad accepts connection, but does not answer commands, when verifying a block |
17:05 |
asciilifeform |
( nuffin 'green' about it, per se, is simply consequence of conventional threading + over9000 locks in there ) |
17:05 |
phf |
well threads work by system level preemption, context switch can happen on instruction, etc. in trb's case the switch happens, there's just nothing to run because everyone's sitting on locks |
17:06 |
asciilifeform |
aha |
17:06 |
asciilifeform |
simply couldn't resist pointing out how author simulated absence of preemption |
17:08 |
phf |
btcbase briefly ran on licensed acl :> |
17:08 |
asciilifeform |
phf: how did that go ? |
17:08 |
asciilifeform |
( or was in prototype stage ? ) |
17:11 |
asciilifeform |
( speaking of commercial lisps, hey thimbronion , didja ever get that 'lispworks' ? ) |
17:11 |
dulapbot |
Logged on 2022-03-18 13:54:46 thimbronion: I'm just gonna try out LispWorks and see how well debugging multiple processes works. Could come in very hande for pest dev. |
17:11 |
phf |
it was the best experience, as always with acl, tight memory usage, lots of useful analysis tools, but i didn't want to ugh use a tool provisioned under a nice fintech contract in association with a project that had a lot of things to say on many subjects :> |
17:14 |
phf |
acl lets you say a lot of things about how the objects should be allocated, and retained, also the annotation system gives you much better integrated tight guarantees. i don't know why but sbcl's optimization techniques still feel like magic often times, and then they fall apart when inputs change. acl expects your commands to be law, and then does runtime checks on boundaries from (speed)(debug) levels |
17:15 |
thimbronion |
asciilifeform: I did |
17:16 |
asciilifeform |
thimbronion: aah? had chance to try ? |
17:17 |
thimbronion |
asciilifeform: I have barely played with it as of yet. |
17:17 |
phf |
there's tons of neat "low level" features. like you can for example say that all the log messages (even though they are fields on structs in my case) should live in the same region, not require gc-ing, should be placed continously by some rule, without dropping down to ffi. that obv makes knuth-morris-pratt search work much better (~~1-2 second faster on eqv machine/code vs sbcl) due to locality |
17:18 |
asciilifeform |
phf: many yrs ago i requisitioned acl ( in same shop where got the bolix... ) but shamefully must admit never had chance to properly put to use. |
17:18 |
phf |
but also presumably the 500mb log messages being eliminated from gc space also makes things faster during gc phase.. |
17:20 |
asciilifeform |
phf: pretty interesting optimizations; very much suggestive of serious use in 'real life' industry |
17:21 |
asciilifeform |
phf: didja end up buying own acl / support service etc ? |
17:23 |
asciilifeform |
( subj, for log readers ) |
| |
~ 15 minutes ~ |
17:38 |
shinohai |
Ukraine yesterday: "We need 7 billion dollars monthly to rebuild." Today: "Bitcoin is banned, lol." |
17:40 |
shinohai |
http://logs.nosuchlabs.com/log/asciilifeform/2022-04-22#1098025 << I just shamelessly code-stole from esthlos and polished for own toolbox, was actually gonna publish new patch so it works on ecl soon. |
17:40 |
dulapbot |
Logged on 2022-04-22 11:06:23 asciilifeform: meanwhile will link to shinohai's alt-vtron. |
17:46 |
asciilifeform |
a |
17:47 |
* |
asciilifeform not tried subj, but iirc shinohai had it working |
17:49 |
asciilifeform |
shinohai: lulzy re ukrs, not so long ago they were begging for btc donations among all else |
17:49 |
shinohai |
Worx fine for me, but naturally "your experience may vary" (tm) |
17:50 |
shinohai |
asciilifeform: yup, likely all donations in "hosted wallets" so means of production now seized. |
17:51 |
shinohai |
"Yo dawg, I heard you like dekulakization ......" |
17:53 |
asciilifeform |
shinohai: through asciilifeform's telescope loox moar like they got call from reich, 'hey yer pushing up exch rate, wtf' |
17:56 |
shinohai |
Probably right since they've spooled up the printing press at 10x speeds for war effort. |
17:56 |
verisimilitude |
I truly hate how e-mail is being destroyed by large corporations, because destruction is all they truly do. |
17:56 |
asciilifeform |
verisimilitude: looong ago destroyed |
17:57 |
asciilifeform |
( try sending from own hosted smtp box to just about anywhere in reich ) |
17:57 |
verisimilitude |
They see some public good, and decide they'd rather force everyone else to suffer by destroying it, purely for their benefit. |
17:57 |
asciilifeform |
silently goes to /dev/null |
17:57 |
verisimilitude |
Oh, not even that, asciilifeform. |
17:58 |
verisimilitude |
Even the smaller large hosts are getting blocked; it doesn't matter. |
17:58 |
asciilifeform |
verisimilitude: the process whereby email was murdered is well-documented (with possible exception of how the perps in fact ~created~ the spam problem which 'justified' the later 'embrace & extinguish' ) |
17:58 |
verisimilitude |
I like how the proof of work method never took off. |
17:58 |
asciilifeform |
verisimilitude: one time asciilifeform witnessed, in commercial shop, microshit's email service swallow mail from ~microshit itself~ |
17:59 |
verisimilitude |
``Noooooooooooo!!!!! Use reputation instead or else we can't send our spam!!!!!!!!!!!!!!!'' |
17:59 |
asciilifeform |
naturally |
18:00 |
verisimilitude |
My personal e-mail server works fine enough; I don't use it much. |
18:00 |
asciilifeform |
worx so long as not need to send, lol |
18:00 |
verisimilitude |
A lot of attractive women want to meet me, for some reason. |
18:00 |
asciilifeform |
princes -- need ransoming, etc |
18:01 |
verisimilitude |
Still, occasionally, I get to send an e-mail to a human who also hosts his, and it's pleasant. |
18:02 |
asciilifeform |
pleasant except for the part where you gotta dig through over9000 'princes' ea. morning |
18:02 |
verisimilitude |
I solve that by checking for e-mail less than once a week. |
18:03 |
asciilifeform |
then simply over9000 x 7 1/wk lol |
18:03 |
verisimilitude |
It sucks how IRC still works the best for some things. |
| |
↖ |
18:03 |
asciilifeform |
verisimilitude: pestnet worx pretty well atm, on thimbronion's proggy |
18:04 |
verisimilitude |
I look forward to using my Pest, if I can help it, but it will have a Spartan interface. |
18:12 |
asciilifeform |
'spartan spreaks with his sword!'(tm)(r) |
18:12 |
asciilifeform |
*speaks |
18:16 |
verisimilitude |
It almost hurts, when an interface flaw about which I'd thought and prevented occurs in something with much more programmers working on it. |
18:16 |
verisimilitude |
Of course, I did notice it requires a holistic approach to prevent. |
18:17 |
verisimilitude |
Perhaps I'd be more humble, if more programmers weren't determined to prove their lack of worth. |
18:21 |
asciilifeform |
verisimilitude: most commercial proggies yer likely to encounter or hear about are braindamaged for economically fundamental reasons |
18:21 |
dulapbot |
Logged on 2020-08-12 13:04:27 asciilifeform: gregorynyssa: it long ago went from 'mistake, like leaded petrol' to deliberate 'job-creating tech' fraud. |
18:23 |
asciilifeform |
http://logs.nosuchlabs.com/log/asciilifeform/2022-04-22#1098183 << worx while the box worx, lol |
18:23 |
dulapbot |
Logged on 2022-04-22 14:02:09 verisimilitude: It sucks how IRC still works the best for some things. |
18:23 |
asciilifeform |
!q uptime |
18:23 |
dulapbot |
asciilifeform: time since my last reconnect : 310d 15h 20m |
18:25 |
* |
asciilifeform even if 0 upstream headaches, is gonna have to clean the fans or sumsuch in it eventually. was at one pt gonna set up an offsite spare, but imho at this pt will say 'pestnet is the spare!' |
| |
↖ |
18:25 |
dulapbot |
Logged on 2022-04-01 16:10:27 asciilifeform: ... for current shitpnoje wireless , is a visible to naked eye 'lolno'. but interestingly enuff asciilifeform's dc in fact delivered 'five nines' service in 2021. i.e. <6min of dead time in yr |
18:29 |
verisimilitude |
Ted Kaczynski never had to deal with this. |
18:29 |
asciilifeform |
verisimilitude: iirc he had to go to town for matches & ammo. so, in a way, did |
18:30 |
verisimilitude |
I meant shitty programs specifically. |
18:32 |
verisimilitude |
How many people need to die before shitty software won't be used in heart monitors and the like? |
18:44 |
verisimilitude |
People don't understand the value of analog devices. Consider a mirror with a camera inside of it. They need to be sat down and told how the mirror is like a computer, but the universe does it in its own little code. |
| |
↖ |
18:45 |
verisimilitude |
The mirror with a camera inside of it is stupid, to clarify. |
| |
~ 23 minutes ~ |
19:08 |
asciilifeform |
verisimilitude: aaapparently is considered a-ok for e.g. boeing's firmware to divide by 0 |
19:08 |
asciilifeform |
so wainot, then, erryone else. |
19:09 |
asciilifeform |
( picture, e.g., what architecture might've been like w/out hammurabi et al ) |
19:09 |
dulapbot |
Logged on 2021-08-09 13:18:25 asciilifeform: mats: structures collapsing ~2y after construction is something i associate with the time of hammurabi ( 'If a builder builds a house for a man and does not make its construction sound, and the house which he has built collapses and causes the death of the owner of the house, the builder shall be put to death...' ) than current day |
19:10 |
asciilifeform |
for that matter, even this -- in process of being 'normalized' as we speak. |
19:10 |
dulapbot |
Logged on 2021-08-09 13:14:07 asciilifeform: punkman: not always. |
19:14 |
asciilifeform |
http://logs.nosuchlabs.com/log/asciilifeform/2022-04-22#1098203 << apropos, and imho sheds light on subj -- turns out that single-seat prop planes still ~100% using carburetted engines w/ 0 electronics. incl. recently-manufactured ones. likely, suspect, because 'it's yer arse', and owner understands, rather than it being corp & insurance wank in abstract, as w/ passenger liners |
19:14 |
dulapbot |
Logged on 2022-04-22 14:43:36 verisimilitude: People don't understand the value of analog devices. Consider a mirror with a camera inside of it. They need to be sat down and told how the mirror is like a computer, but the universe does it in its own little code. |
19:16 |
* |
asciilifeform vaguely recalls an article by signpost where he was flying in one and described this |
19:21 |
asciilifeform |
( meanwhile you prolly can't buy a flashlight w/out a computer. and crashes. Because Reasons ) |
| |
↖ |
19:21 |
dulapbot |
Logged on 2020-05-10 22:54:23 asciilifeform: i've found flashlights w/ 8bit micros in'em. |
| |
~ 1 hours 2 minutes ~ |
20:24 |
signpost |
yep, in shitworld, the more money being allocated to a problem, the more brittle the solution may be. |
20:27 |
* |
signpost currently grunting through a lul where for some reason, after bootstrapping gnat, the compiler cannot find the "Ada" package, thus can't rebuild itself. |
20:28 |
signpost |
can't help but see an analogy here. |
20:30 |
signpost |
every distro, every company, whoever touched gcc, there was always a process of digestion before the shitwad was usable. |
20:31 |
* |
signpost bothering to not just "eh, I got this thing to build one time, here's a signed bin which will later be a religious artefact" but actually produce an item that can be bootstrapped on several systems. |
20:33 |
mats |
https://archive.ph/3vces |
| |
↖ |
20:34 |
signpost |
heh, onoes, how will the russian crypto industry access capital. |
| |
~ 42 minutes ~ |
21:17 |
shinohai |
Link shared with me from a twatter denizen on why trb ain't a full Bitcoin node: https://archive.ph/meMzf |
21:17 |
shinohai |
many keks were had. |
| |
~ 1 hours 42 minutes ~ |
23:00 |
asciilifeform |
shinohai: lulzy. (aand see also) |
23:00 |
dulapbot |
(trilema) 2015-12-19 asciilifeform: http://log.bitcoin-assets.com/?date=19-12-2015#1348050 << you know what's great? rat poison STILL WORKS even if you don't explain to the rats exactly ~how~ ! |
23:03 |
asciilifeform |
http://logs.nosuchlabs.com/log/asciilifeform/2022-04-22#1098221 << lol, so much for this |
23:03 |
dulapbot |
Logged on 2022-04-22 16:32:32 mats: https://archive.ph/3vces |
23:03 |
dulapbot |
Logged on 2022-01-25 13:03:51 mats: my initial arg was that cz/binance is a sovereign, and i think that's borne out of the facts |